skip to main content
DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Deep Learning for Semantic Segmentation of Defects in Advanced STEM Images of Steels

Abstract

Crystalline materials exhibit long-range ordered lattice unit, within which resides nonperiodic structural features called defects. These crystallographic defects play a vital role in determining the physical and mechanical properties of a wide range of material systems. While computer vision has demonstrated success in recognizing feature patterns in images with well defined contrast, automated identification of nanometer scale crystallographic defects in electron micrographs governed by complex contrast mechanism is still a challenging task. Here, building upon an advanced defect imaging mode that offers high feature clarity, we introduce DefectSegNet - a new convolutional neural network (CNN) architecture that performs semantic segmentation of three common crystallographic defects in structural alloys: dislocation lines, precipitates and voids. Results from supervised training of a small set of high-quality defect images of steels show high pixel-wise accuracy across all three types of defects: 91.60 ± 1.77% on dislocations, 93.39 ± 1.00% on precipitates, and 98.85 ± 0.56% on voids. We discuss the sources of uncertainties in CNN prediction, and the training data in terms of feature density, representation and homogeneity and their effects on deep learning performance. Further defect quantification using DefectSegNet prediction outperforms human expert average, presenting a promising new workflow for fast and statisticallymore » meaningful quantification of materials defects.« less

Authors:
 [1];  [2];  [3]; ORCiD logo [1];  [4]; ORCiD logo [5]
  1. Pacific Northwest National Lab. (PNNL), Richland, WA (United States)
  2. Western Washington Univ., Bellingham, WA (United States)
  3. Univ. of Connecticut, Storrs, CT (United States)
  4. Western Washington Univ., Bellingham, WA (United States); Pacific Northwest National Lab. (PNNL), Richland, WA (United States)
  5. Pacific Northwest National Lab. (PNNL), Richland, WA (United States); Univ. of Connecticut, Storrs, CT (United States)
Publication Date:
Research Org.:
Pacific Northwest National Lab. (PNNL), Richland, WA (United States)
Sponsoring Org.:
USDOE Office of Science (SC), Fusion Energy Sciences (FES) (SC-24)
OSTI Identifier:
1559992
Report Number(s):
PNNL-SA-146546
Journal ID: ISSN 2045-2322
Grant/Contract Number:  
AC05-76RL01830
Resource Type:
Accepted Manuscript
Journal Name:
Scientific Reports
Additional Journal Information:
Journal Volume: 9; Journal Issue: 1; Journal ID: ISSN 2045-2322
Publisher:
Nature Publishing Group
Country of Publication:
United States
Language:
English
Subject:
machine leaning; STEM Imaging; convolutional neural network; semantic segmentation

Citation Formats

Roberts, Graham, Haile, Simon, Sainju, Rajat, Edwards, Danny J., Hutchinson, Brian J., and Zhu, Yuanyuan. Deep Learning for Semantic Segmentation of Defects in Advanced STEM Images of Steels. United States: N. p., 2019. Web. doi:10.1038/s41598-019-49105-0.
Roberts, Graham, Haile, Simon, Sainju, Rajat, Edwards, Danny J., Hutchinson, Brian J., & Zhu, Yuanyuan. Deep Learning for Semantic Segmentation of Defects in Advanced STEM Images of Steels. United States. doi:10.1038/s41598-019-49105-0.
Roberts, Graham, Haile, Simon, Sainju, Rajat, Edwards, Danny J., Hutchinson, Brian J., and Zhu, Yuanyuan. Wed . "Deep Learning for Semantic Segmentation of Defects in Advanced STEM Images of Steels". United States. doi:10.1038/s41598-019-49105-0. https://www.osti.gov/servlets/purl/1559992.
@article{osti_1559992,
title = {Deep Learning for Semantic Segmentation of Defects in Advanced STEM Images of Steels},
author = {Roberts, Graham and Haile, Simon and Sainju, Rajat and Edwards, Danny J. and Hutchinson, Brian J. and Zhu, Yuanyuan},
abstractNote = {Crystalline materials exhibit long-range ordered lattice unit, within which resides nonperiodic structural features called defects. These crystallographic defects play a vital role in determining the physical and mechanical properties of a wide range of material systems. While computer vision has demonstrated success in recognizing feature patterns in images with well defined contrast, automated identification of nanometer scale crystallographic defects in electron micrographs governed by complex contrast mechanism is still a challenging task. Here, building upon an advanced defect imaging mode that offers high feature clarity, we introduce DefectSegNet - a new convolutional neural network (CNN) architecture that performs semantic segmentation of three common crystallographic defects in structural alloys: dislocation lines, precipitates and voids. Results from supervised training of a small set of high-quality defect images of steels show high pixel-wise accuracy across all three types of defects: 91.60 ± 1.77% on dislocations, 93.39 ± 1.00% on precipitates, and 98.85 ± 0.56% on voids. We discuss the sources of uncertainties in CNN prediction, and the training data in terms of feature density, representation and homogeneity and their effects on deep learning performance. Further defect quantification using DefectSegNet prediction outperforms human expert average, presenting a promising new workflow for fast and statistically meaningful quantification of materials defects.},
doi = {10.1038/s41598-019-49105-0},
journal = {Scientific Reports},
number = 1,
volume = 9,
place = {United States},
year = {2019},
month = {9}
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record

Save / Share: